Concept: Regression analysis
- Proceedings of the National Academy of Sciences of the United States of America
- Published over 2 years ago
The most widely used task functional magnetic resonance imaging (fMRI) analyses use parametric statistical methods that depend on a variety of assumptions. In this work, we use real resting-state data and a total of 3 million random task group analyses to compute empirical familywise error rates for the fMRI software packages SPM, FSL, and AFNI, as well as a nonparametric permutation method. For a nominal familywise error rate of 5%, the parametric statistical methods are shown to be conservative for voxelwise inference and invalid for clusterwise inference. Our results suggest that the principal cause of the invalid cluster inferences is spatial autocorrelation functions that do not follow the assumed Gaussian shape. By comparison, the nonparametric permutation test is found to produce nominal results for voxelwise as well as clusterwise inference. These findings speak to the need of validating the statistical methods being used in the field of neuroimaging.
Despite the many number of studies examining workaholism, large-scale studies have been lacking. The present study utilized an open web-based cross-sectional survey assessing symptoms of psychiatric disorders and workaholism among 16,426 workers (Mage = 37.3 years, SD = 11.4, range = 16-75 years). Participants were administered the Adult ADHD Self-Report Scale, the Obsession-Compulsive Inventory-Revised, the Hospital Anxiety and Depression Scale, and the Bergen Work Addiction Scale, along with additional questions examining demographic and work-related variables. Correlations between workaholism and all psychiatric disorder symptoms were positive and significant. Workaholism comprised the dependent variable in a three-step linear multiple hierarchical regression analysis. Basic demographics (age, gender, relationship status, and education) explained 1.2% of the variance in workaholism, whereas work demographics (work status, position, sector, and annual income) explained an additional 5.4% of the variance. Age (inversely) and managerial positions (positively) were of most importance. The psychiatric symptoms (ADHD, OCD, anxiety, and depression) explained 17.0% of the variance. ADHD and anxiety contributed considerably. The prevalence rate of workaholism status was 7.8% of the present sample. In an adjusted logistic regression analysis, all psychiatric symptoms were positively associated with being a workaholic. The independent variables explained between 6.1% and 14.4% in total of the variance in workaholism cases. Although most effect sizes were relatively small, the study’s findings expand our understanding of possible psychiatric predictors of workaholism, and particularly shed new insight into the reality of adult ADHD in work life. The study’s implications, strengths, and shortcomings are also discussed.
The exact timing, route, and process of the initial peopling of the Americas remains uncertain despite much research. Archaeological evidence indicates the presence of humans as far as southern Chile by 14.6 thousand years ago (ka), shortly after the Pleistocene ice sheets blocking access from eastern Beringia began to retreat. Genetic estimates of the timing and route of entry have been constrained by the lack of suitable calibration points and low genetic diversity of Native Americans. We sequenced 92 whole mitochondrial genomes from pre-Columbian South American skeletons dating from 8.6 to 0.5 ka, allowing a detailed, temporally calibrated reconstruction of the peopling of the Americas in a Bayesian coalescent analysis. The data suggest that a small population entered the Americas via a coastal route around 16.0 ka, following previous isolation in eastern Beringia for ~2.4 to 9 thousand years after separation from eastern Siberian populations. Following a rapid movement throughout the Americas, limited gene flow in South America resulted in a marked phylogeographic structure of populations, which persisted through time. All of the ancient mitochondrial lineages detected in this study were absent from modern data sets, suggesting a high extinction rate. To investigate this further, we applied a novel principal components multiple logistic regression test to Bayesian serial coalescent simulations. The analysis supported a scenario in which European colonization caused a substantial loss of pre-Columbian lineages.
Risk sharing arrangements between hospitals and payers together with penalties imposed by the Centers for Medicare and Medicaid (CMS) are driving an interest in decreasing early readmissions. There are a number of published risk models predicting 30 day readmissions for particular patient populations, however they often exhibit poor predictive performance and would be unsuitable for use in a clinical setting. In this work we describe and compare several predictive models, some of which have never been applied to this task and which outperform the regression methods that are typically applied in the healthcare literature. In addition, we apply methods from deep learning to the five conditions CMS is using to penalize hospitals, and offer a simple framework for determining which conditions are most cost effective to target.
Fructose has been implicated in the pathogenesis of obesity and type 2 diabetes. In contrast to glucose, CNS delivery of fructose in rodents promotes feeding behavior. However, because circulating plasma fructose levels are exceedingly low, it remains unclear to what extent fructose crosses the blood-brain barrier to exert CNS effects. To determine whether fructose can be endogenously generated from glucose via the polyol pathway (glucose → sorbitol → fructose) in human brain, 8 healthy subjects (4 women/4 men; age, 28.8 ± 6.2 years; BMI, 23.4 ± 2.6; HbA1C, 4.9% ± 0.2%) underwent (1)H magnetic resonance spectroscopy scanning to measure intracerebral glucose and fructose levels during a 4-hour hyperglycemic clamp (plasma glucose, 220 mg/dl). Using mixed-effects regression model analysis, intracerebral glucose rose significantly over time and differed from baseline at 20 to 230 minutes. Intracerebral fructose levels also rose over time, differing from baseline at 30 to 230 minutes. The changes in intracerebral fructose were related to changes in intracerebral glucose but not to plasma fructose levels. Our findings suggest that the polyol pathway contributes to endogenous CNS production of fructose and that the effects of fructose in the CNS may extend beyond its direct dietary consumption.
BACKGROUND: The aim of this study is to compare the odds of postpartum haemorrhage among women who opt for home birth against the odds of postpartum haemorrhage for those who plan a hospital birth. It is an observational study involving secondary analysis of maternity records, using binary logistic regression modelling. The data relate to pregnancies that received maternity care from one of fifteen hospitals in the former North West Thames Regional Health Authority Area in England, and which resulted in a live or stillbirth in the years 1988–2000 inclusive, excluding ‘high-risk’ pregnancies, unplanned home births, pre-term births, elective Caesareans and medical inductions. RESULTS: Even after adjustment for known confounders such as parity, the odds of postpartum haemorrhage (>=1000ml of blood lost) are significantly higher if a hospital birth is intended than if a home birth is intended (odds ratio 2.5, 95% confidence interval 1.7 to 3.8). The ‘home birth’ group included women who were transferred to hospital during labour or shortly after birth. CONCLUSIONS: Women and their partners should be advised that the risk of PPH is higher among births planned to take place in hospital compared to births planned to take place at home, but that further research is needed to understand (a) whether the same pattern applies to the more life-threatening categories of PPH, and (b) why hospital birth is associated with increased odds of PPH. If it is due to the way in which labour is managed in hospital, changes should be made to practices which compromise the safety of labouring women.
Correctly assessing a scientist’s past research impact and potential for future impact is key in recruitment decisions and other evaluation processes. While a candidate’s future impact is the main concern for these decisions, most measures only quantify the impact of previous work. Recently, it has been argued that linear regression models are capable of predicting a scientist’s future impact. By applying that future impact model to 762 careers drawn from three disciplines: physics, biology, and mathematics, we identify a number of subtle, but critical, flaws in current models. Specifically, cumulative non-decreasing measures like the h-index contain intrinsic autocorrelation, resulting in significant overestimation of their “predictive power”. Moreover, the predictive power of these models depend heavily upon scientists' career age, producing least accurate estimates for young researchers. Our results place in doubt the suitability of such models, and indicate further investigation is required before they can be used in recruiting decisions.
Studies have shown links between educational outcomes such as letter grades, test scores, or other measures of academic achievement, and health-related behaviors (1-4). However, as reported in a 2013 systematic review, many of these studies have used samples that are not nationally representative, and quite a few studies are now at least 2 decades old (1). To update the relevant data, CDC analyzed results from the 2015 national Youth Risk Behavior Survey (YRBS), a biennial, cross-sectional, school-based survey measuring health-related behaviors among U.S. students in grades 9-12. Analyses assessed relationships between academic achievement (i.e., self-reported letter grades in school) and 30 health-related behaviors (categorized as dietary behaviors, physical activity, sedentary behaviors, substance use, sexual risk behaviors, violence-related behaviors, and suicide-related behaviors) that contribute to leading causes of morbidity and mortality among adolescents in the United States (5). Logistic regression models controlling for sex, race/ethnicity, and grade in school found that students who earned mostly A’s, mostly B’s, or mostly C’s had statistically significantly higher prevalence estimates for most protective health-related behaviors and significantly lower prevalence estimates for most health-related risk behaviors than did students with mostly D’s/F’s. These findings highlight the link between health-related behaviors and education outcomes, suggesting that education and public health professionals can find their respective education and health improvement goals to be mutually beneficial. Education and public health professionals might benefit from collaborating to achieve both improved education and health outcomes for youths.
Thousands of lives are lost every year in developing countries for failing to detect epidemics early because of the lack of real-time disease surveillance data. We present results from a large-scale deployment of a telephone triage service as a basis for dengue forecasting in Pakistan. Our system uses statistical analysis of dengue-related phone calls to accurately forecast suspected dengue cases 2 to 3 weeks ahead of time at a subcity level (correlation of up to 0.93). Our system has been operational at scale in Pakistan for the past 3 years and has received more than 300,000 phone calls. The predictions from our system are widely disseminated to public health officials and form a critical part of active government strategies for dengue containment. Our work is the first to demonstrate, with significant empirical evidence, that an accurate, location-specific disease forecasting system can be built using analysis of call volume data from a public health hotline.
OBJECTIVE The possible interaction of serum 25-hydroxyvitamin D [25(OH)D] and obesity in regard to type 2 diabetes and insulin resistance has not been well studied. To explore the effect modification of obesity on the association between 25(OH)D and insulin resistance/type 2 diabetes, data were examined from a nationally representative sample. RESEARCH DESIGN AND METHODS The analytic sample for the type 2 diabetes analysis (n = 12,900) was limited to participants from the National Health and Nutrition Examination Survey (NHANES) 2001-2006 over 20 years of age. Participants >20 years of age assigned to the morning session and free of diabetes were limited to the insulin resistance analysis (n = 5,806). Multiplicative interaction was assessed through a cross-product interaction term in a multiple logistic regression model. The presence of additive interaction between insufficient 25(OH)D and obesity (indicated by BMI or waist circumference) was evaluated by calculation of the relative excess risk due to interaction (RERI) and attributable proportion due to interaction (AP). RESULTS There was no multiplicative interaction of insufficient 25(OH)D and obesity on type 2 diabetes or insulin resistance. Furthermore, none of the RERI or AP values were statistically significant in the diabetes analysis. However, there was strong additive interaction between abdominal obesity and insufficient 25(OH)D (RERI 6.45 [95% CI 1.03-11.52]) in regard to insulin resistance. In addition, 47% of the increased odds of insulin resistance can be explained by interaction between insufficient 25(OH)D and high BMI (AP 0.47 [95% CI 0.08-0.87]). CONCLUSIONS Within a cross-sectional, nationally representative sample, abdominal obesity and insufficient 25(OH)D interact to synergistically influence the risk of insulin resistance.